Professional Documents
Culture Documents
Big Data Marketing
Big Data Marketing
TABLE OF CONTENTS
1. FOREWORD 4
1.1 MARKETING 5
Page 2
BIG DATA MARKETING
4. LOYALTY PROGRAMS 74
3.3.4 OMNICHANEL 93
5. ANNEXES 101
6. BIBLIOGRAPHY 110
Page 3
BIG DATA MARKETING
0. FOREWORD
In today’s competitive world, one of our main objectives must be to retain our
customers and keep them completely satisfied since the company would not exist
without loyal customers.
Throughout history, companies have used marketing to take some actions and offer
promotions in order to generate more sales and increase customer satisfaction and
loyalty. However, this is not enough since personalised attention has become essential
to make a difference.
Promotions to satisfy our customers are not enough now, but these promotions should
be as specific and personalised as possible, so that they adapt completely to our
customers, considering their opinions, tastes, preferences and expectations.
This manual will address marketing from the perspective of big data to see how to use
all the data and information of our customers and the environment to plan a strategy or
marketing strategies adapted to each customer.
We cannot forget that customers are essential for the company, so planning actions and
personalised strategies will ensure our continuity, as well as the loyalty of our best
customers.
Page 4
BIG DATA MARKETING
1.1 MARKETING
As the name of the manual suggests, Big Data marketing will be addressed here.
However, before focusing on this, it is particularly important to summarise what
marketing is as this will help us to better understand the topics discussed later.
Marketing can be defined as the field that analyses market and consumer behaviour.
The primary purpose of analysing the commercial management of companies is to
attract and retain customers by satisfying their needs.
Marketing is often related only with sales. This is wrong since marketing must be seen
as something related to satisfying customer needs. In addition to sales and advertising,
marketing should also consider the following aspects:
Share these offers and help the audience know them effectively
When we talk about marketing, the following basic concepts must be taken into
account:
▪ Needs, desires and demands: the needs of people are the starting point of
marketing. These needs are not created by society or marketing specialists, but
exist in humans. These needs may be physical, such as eating or feeling safe;
social, such as feeling accepted and part of a group, and individual, such as
personal self-realisation, among others. So, we should detect what people want
and give it to them.
Page 5
BIG DATA MARKETING
Desires are defined as the lack of something specific that satisfies basic needs.
For example, a person is cold and wants a fur coat, is hungry and wants a steak,
or needs to feel important and wants an Armani suit.
The desires are obviously related to the society in which people live. A desire is
not satisfied in the same way in the United States as in Indonesia.
▪ Products and services: Individuals meet their needs and desires with products
and services.
▪ Value, cost and satisfaction: when consumers decide on a product from among
the huge offer, they make their decision based on the net value expectations of
the different offers.
Value expectations refer to the difference between the positive values and the
negative values that consumers expect when they buy a product.
This value is an integral element of the level of customer satisfaction, and affects
their future behaviour. This refers to the perception of the values received when
the product has been purchased. That is, it involves comparing those values that
they expected to find with the values they have found.
Page 6
BIG DATA MARKETING
For the exchange to take place, the following conditions must be met:
The exchange should be seen as a process rather than as an event. It is said that
two people are in a process of exchange if they are negotiating to reach an
agreement. If this agreement is reached, what is known as a transaction has
taken place.
Thus, the size of the market depends on the number of people who have a
specific need, have resources that the other party wants and wish to exchange
these resources for what they need.
Page 7
BIG DATA MARKETING
For such exchanges to be possible, two parties are needed. If one of these two
parties is more active than the other in the search for exchange, it is called
exchange finder, while the other is called receiver.
It can be said that the exchange finder is a company that works in a market of
end users in competition with other companies. These companies send their
messages and products to the end users directly or through an intermediary. Its
effectiveness is influenced by its suppliers, as well as by different environmental
forces. These forces can be: demographic, economic, physical, technological,
political, legal and socio-cultural, among others.
SUPPLIERS
COMPETITION COMPANY
INTERMEDIARIES CUSTOMERS
USER MARKET
Page 8
BIG DATA MARKETING
Given the elements that make up marketing, it is also interesting to briefly introduce
what is known as marketing management.
This management takes place when at least one of the parties to a potential exchange
thinks about the objectives and means with which to achieve desired responses from
other parties.
Demand management is responsible for influencing the level, timing and composition
of the demand to help the organisation achieve its objectives. It can be said that
marketing management is essentially based on demand management.
Every company frequently establishes the desired level of transactions with a target
audience. The demand reaches that level sometimes, but it may be higher, lower or
there may even be no demand on other occasions.
Page 9
BIG DATA MARKETING
management has been aimed at creating relationships with current customers, in order
to get loyal and long-term customers.
Due to the current competitive business environment, this customer retention and
loyalty is very important so personalised service becomes more important in order to
prevent our customers leaving for the competition. In addition to this, attracting new
customers involves a really high cost for the company. It is said that it is five times more
expensive to attract new customers than to keep a customer satisfied.
In addition to the high cost of attracting a new customer, we must also consider what
losing a regular customer means, as this involves losing all the purchases that customer
could make in our company. This has created two new concepts: the customer lifetime
value and the cost of losing a customer. The latter is really important, since we not only
lose that customer, but the reason for this loss can mean the loss of other customers.
For example, if one of our customers is dissatisfied with the customer service received
or with the product or service purchased, they can share this dissatisfaction with their
friends or relatives and influence them to leave us if they were our customers, or not to
become our customers.
Social media must also be taken into account, since if a dissatisfied customer shares their
experience on social media, the number of recipients of this bad experience will be
higher than if they merely say it to their group of friends. That is why demand
management and long relationships with our customers are fundamental.
Regarding these guidelines, there are five approaches that companies or organisations
can use to develop their exchange activities. These approaches are production, product,
sales, marketing and social marketing.
Page 10
BIG DATA MARKETING
These five approaches will be briefly introduced, and strategic planning and the
marketing process will be discussed later.
▪ Production approach
The oldest approach in exchange management, this approach argues that consumers
will favour the products that are very available and low cost. The managers of the
companies that follow this approach will use their efforts to achieve economies of scale
and wide distribution. This philosophy is focused on making products rather than on
getting customers.
- In countries where the demand for a product outstrips supply. This generally
happens in third world countries, where consumers are more interested in
getting the product than in its benefits, so producers direct their efforts at
increasing production by offering products at a low price.
- When the cost of the product is high and it must be reduced by improving
productivity to increase the market.
To summarise, this approach is about producing a lot to ensure low production costs, as
well as a low sales price.
Page 11
BIG DATA MARKETING
▪ Product approach
This second approach argues that consumers will favour products that offer the best
quality or the best results. In this case, the efforts will be put in manufacturing good
products and improving them over time.
The managers of the companies or organisations assume that buyers admire and want
well-made and quality products, valuing their advantages and benefits.
The companies that follow this approach often pay more attention to the product than
to the market and the demand. This can be negative, since they do not consider the
demand when creating and improving their products, which means that they may create
useless products. This is known as marketing myopia, which refers to the focus on the
product and not on the need.
▪ Sales approach
As its name suggests, this approach focuses on sales exclusively. This means that the
company must carry out aggressive sales and promotion policies to sell as much as
possible. It is basically about selling what they produce, not producing what they can
sell.
This approach is mainly used in unwanted goods, that is, those goods that the consumer
does not intend to buy or does not usually buy. However, it is also used with desired
products or with products of non-profit areas.
It is also a widely used approach when companies have an excess of productive capacity
and want to get rid of the available stock.
Page 12
BIG DATA MARKETING
▪ Marketing approach
The marketing approach is the opposite of the approaches presented so far. It says that
to achieve the objectives, companies must identify the needs and wants of their
customers and target audiences, in order to effectively meet their needs. It can be said
that:
- The marketing approach tries to satisfy the customer's needs through the
product and the set of benefits associated with its delivery and consumption.
This approach is based on four essential pillars: market definition, customer orientation,
marketing coordination and profitability.
Market definition
It is impossible for a company to operate in all markets and meet all needs. Therefore,
each company must define its target audience carefully since actions will be taken to
satisfy the needs of a group depending on the type of audience and its characteristics.
For example, a travel agency must determine what type of customer they want to
address, as this will determine the product they offer and the way to offer it. Thus,
organising luxury trips for newlyweds will not be the same as organising trips for
university students or retired people.
Customer orientation
Once our target audience is defined, we should define their needs from their point of
view, not from the point of view of the company. This is not always easy since we must
interpret the reality of customers.
Page 13
BIG DATA MARKETING
- Real needs à the customer wants a car whose operating costs, not its purchase
price, are low.
- Unstated needs à the customer wants good treatment and service from the
selling company.
- Delight needs à the customer receives a gift or additional service for the
purchase of the car.
Each product has a number of aspects. The marketing department should investigate
them to get to know those aspects that our target audience most values in order to be
able to offer products that fully or nearly meet their needs. We cannot forget that the
main objective of this approach is to sell through the satisfaction of the customers’
needs since a satisfied customer:
- Speaks positively with other people about the company and the products and
services.
Page 14
BIG DATA MARKETING
As has already been said, customer satisfaction is essential since, as a satisfied customer
will talk about the company and its products and services positively, while a dissatisfied
customer will do so negatively and this will contribute to harming the reputation and
image of the company.
For this, it is very important for companies to regularly measure the degree of
satisfaction of their customers, considering not only the good, but also those negative
aspects such as complaints or claims, as this will help us improve.
Marketing coordination
For this, it is essential that the company not only carries out external marketing actions,
but also internal marketing actions. Internal marketing refers to hiring, training and
motivating staff to assist customers adequately. Thus, internal marketing must precede
external marketing, as it is of no use to promise excellent service to customers when the
company is not prepared for it internally.
Page 15
BIG DATA MARKETING
Profitability
One of the objectives of the marketing approach is to help organisations achieve their
objectives, which are usually related to maximize their profits.
Therefore, the marketing department should not only work on the goal to satisfy
customers, but also to operate profitability.
Further discussing what was introduced in the marketing approach, the social marketing
approach considers society in addition to customer satisfaction and the profitability of
the company. It takes into account the public interest of society when making decisions.
Linked with this approach, Corporate Social Responsibility should be mentioned. This is
a way of managing companies based on the management of the impacts that their
activity generates on their customers, employees, shareholders, the environment and
society in general. It achieves a balance between economic growth, social welfare,
natural resources and the environment.
This balance between business and society is essential for business, because if the
company is an active part in solving the challenges of society, this will result in:
- Higher productivity through better conditions for the internal customer that
leads to better talent retention and, consequently, lower turnover rates.
- Loyalty to the customer and satisfaction of their needs, first by providing a place
where they can make their needs and complaints known. In addition to quality
and price, customers have started to demand information about production
conditions and product certifications, among others.
Page 16
BIG DATA MARKETING
In addition to being socially responsible with the environment, this will be achieved if
the company is transparent with the information it offers society in relation to its
practices and management. It must also respond to the expectations of society, as well
as adapt to the market and the needs of its different stakeholders.
Using the definition of corporate social responsibility as a reference, it can be said that
the focus on social marketing is on being socially responsible, so that it not only provides
benefits for the company and customers, but also for society and the environment.
Given these considerations about traditional marketing, the following chapter deals
with what is known as big data marketing.
As seen in other manuals, companies have a large amount of data about their customers
and consumers. These data can be extracted from their purchases or surveys, or data
may come from comments and publications of these consumers on their social media.
All this gives us valuable and useful information about our customers, which allows our
marketing strategies to be directed to a customer in a personalised way instead of being
addressed to a group of clients. That is what big data marketing is about. It uses all this
information to exclusively address a customer and offer them what they need. It avoids
generic campaigns that do not have the expected results.
As seen in the first part of this chapter, marketing aims to achieve customer satisfaction
by implementing various actions and strategies.
Page 17
BIG DATA MARKETING
In Big Data marketing, we try to take advantage of the data and information we have
about our customers to design fully customised marketing strategies and promotions.
The two main elements that must be taken into account to successfully work with big
data marketing are:
- Strategy: even if our company has a big data approach and deals with customers
in a personalised way. This does not mean that the brand is adapted to each
customer. For this, it is necessary to define what kind of customer we are
interested in and in what way we are going to approach them. The following
actions are involved:
o Define the analytical strategy, that is, establish what data will be used
and stored, and establish what attributes are key to the business.
Page 18
BIG DATA MARKETING
The customer becomes the main element of big data marketing, so marketing must be
adapted to the customer's life cycle.
As we already know, customer analytics is the tool used by companies to analyse all their
customers’ data. It makes a valuable contribution, and that is why customer analytics
becomes an essential tool. In an environment where we know our customers very well,
their management will change.
For this, we must see the customer not as a person who has contact with the brand, but
as a person who has a life with the brand. That is why the concept of the customer life
cycle is important. It consists of the following six stages:
- Buy
- Use
- Ask
- Become a fan
We should manage this life cycle in the best possible way so that the customer that
discovers and explores our brand can buy and become a fan, that is, a very valuable
customer.
The following table shows the previous stages, their business objectives and the
analytical methods that will be used at each stage.
Page 19
BIG DATA MARKETING
Page 20
BIG DATA MARKETING
In addition to these stages of the customer life cycle, it is also important that the
customer management carried out by big data marketing considers what is known as
the customer journey, already seen in the module about Customer Analytics.
The customer journey is a tool of Design Thinking that allows us to capture on a map
each of the stages, interactions, channels and elements customers go through when
interacting with a company, from buying products online to accessing customer service
on the phone to airing grievances on social media.
It is important to highlight that a customer journey does not have to start when the
person starts to be our customer and end when the relationship ends, but that map can
be centralised in a specific stage, such as the purchase process or the attention by the
customer service department in relation to the resolution of a problem. In this way, we
can establish the start and end point depending on our needs.
The following aspects must be taken into account for these maps:
- Understand the stages of the link: as the process is from the customer’s
perspective, we can understand how the customer feels about each of
the interactions with the company, service or product. We must basically
observe all these interactions from the customer’s perspective in order
to determine how they feel at each moment and take advantage of the
results to improve our strategies.
Page 21
BIG DATA MARKETING
Page 22
BIG DATA MARKETING
Mapping this journey will help us to better understand the motivations and satisfaction
of our customers. This will allow us to carry out proper personalised management and
encourage our customer to go through all the stages of the life cycle and become a fan
and a valuable customer.
This will only be achieved with a very good previous analysis of our customer and with
the implementation of actions and marketing promotions related to the results obtained
from our analysis.
Given these considerations, the tools that should be used in big data marketing will now
be discussed.
In addition to the analysis of customers, tools are fundamental for successful big data
marketing.
- Email marketing
- Mobile marketing
- Social marketing
These practices are discussed below to understand how they work and how they should
be used.
Page 23
BIG DATA MARKETING
▪ Email marketing
As its name suggests, it is a direct marketing tool that uses email for advertising and
promotional campaigns.
Over the years and thanks to the rise of the internet, this tool has become very popular.
It has almost completely replaced promotions and advertisements sent by post due to
the fact that email costs less, its reception is immediate and its result can be measured.
An email marketing campaign or strategy may seem as simple as creating an email and
sending it, but it is not as easy as that. This action involves good planning and monitoring
of many processes.
The following three stages must be considered to carry out an email marketing strategy:
Planning
In this first stage, the goals of the campaign must be set. For this, it will be necessary to
establish the needs to be met in the short, medium and long term.
Page 24
BIG DATA MARKETING
Execution
Results measurement
We should not only send the email to our customers, but every action on email
marketing should help us to better know these customers.
The results that this campaign gives us must be measured to use them in future
campaigns.
The main metrics of email marketing that must be considered are those that measure
the effectiveness of the campaign in its different stages:
- Deliverability
- Open rate
- Click rate
- Conversion rate
Big data and customer analytics have been a step further in the implementation of email
marketing campaigns. The following elements must be taken into account when using
email marketing:
Page 25
BIG DATA MARKETING
Target
Thanks to big data and customer analytics, we know our customers better, and this helps
us when establishing our target. This will also help us avoid mass emails and we will only
send that email to the customers who, based on the data we have, may be interested.
We cannot forget that one of the main objectives of marketing is to achieve lasting
relationships with customers, and it is essential to communicate only with the right
people to achieve that.
If we know that some information about certain products may not interest any of our
customers, we should not send them this information as this can tire them, which, in
turn, can lead them to unsubscribe from our database and stop being our customers.
Thanks to big data, we can move from segments to micro-segments, which allow us to
design a personalised strategy because we can decide who to send our emails to.
Content personalisation
Page 26
BIG DATA MARKETING
When we personalise the messages to our customers, we will have to find a point in
which the resources of all type used to personalise the communication are similar to the
additional gain that the results from this personalisation provide. That is, in economic
terms, this personalisation must be profitable, so that the benefits are higher than the
costs.
Once the audience that will receive the email as well as the content of the email is
established, the key moment of the campaign arrives: the delivery. The moment of truth
in our campaign will be when our email is in the inbox. If this moment is right, the
opening rates will skyrocket. Otherwise, the chances of an unread email increase
considerably.
We must take into account that a quarter of the emails received are opened during the
first hour. This percentage falls as time go by. 24 hours after its reception, less than 1%
of these emails are opened.
- According to statistics, the best day and time to send our email marketing
campaigns is on Thursday from 8 to 9 am, followed by Wednesday from 8 to 10
am.
However, we do not need to stick to these principles thanks to big data, as we can define
the best shipping time based on the deep and detailed analysis of our customers.
Page 27
BIG DATA MARKETING
- Create an open frequency variable at the user level that allows us to determine
the frequency of sending an email per user.
Staging
When sending our mails, it is important to distinguish between two types of emails:
transactional emails and brand emails.
- Transactional emails: these mails are aimed at sales. With big data, we can
determine which products may interest our customers, what offers they prefer,
how to get a higher average in the average ticket, etc. We will use these emails
when we want to increase our sales.
- Brand emails: these emails are personalised so that they communicate only what
the customer is interested in about the brand, contributing to improving our
image and reputation.
In addition to selling or improving our brand image and reputation, an email marketing
action must help the company to collect data that may be very useful for future
campaigns.
In order to obtain more data through this type of campaign, we must send an email in
which each part that the receiver may click can be identified. The easiest way to do this
is to categorise the clicks by variables such as the category or the type of action, since
this will facilitate the analysis. Thanks to the categorised clicks, the data model can
include where the user clicks.
For example, imagine that we have a clothing brand that sells women’s clothing, men’s
clothing, home accessories, and children's clothing. If we send an email to our customers
to tell them that sales have started and the customer can access those four sections in
the email, we will be able to identify which section interests each customer, and this will
allow us to be more precise in our future mails. That means we will not include the
Page 28
BIG DATA MARKETING
information they did not click, but we will provide more information about what they
clicked.
▪ Mobile marketing
Mobile marketing is another of the practices of big data marketing. According to the
Mobile Marketing Association, it is defined as a set of practices that enables
organisations to communicate and engage with their audience in an interactive and
relevant manner through any mobile device or network. This not only refers to mobile
phones, but also to any other mobile device, such as tablets.
Page 29
BIG DATA MARKETING
This marketing method is different from traditional marketing as it has a wide range of
means and possibilities to reach users. These are:
- Location: mobile devices have a GPS system through which we can know the
geographic location of our customers. This can be used to show ads according to
the location, show relevant content or offer promotions or coupons.
- Apps: our products and services can be shown to customers through apps, and
they do not need to be physically in the store or in front of the computer. This
means they can buy our product from anywhere.
- Mobile ads: the origin of these ads is in traditional banners. These ads pop up to
the user when they see content on their mobile device and are related to their
preferences. These ads are in video format and allow the user to interact with
them.
- SMS and MMS: although this practice is less popular now due to the emergence
of other practices, promotions through text messages are a quick and direct way
to send the information to our customers. Currently they complement
multichannel campaigns.
- Mobile searches: search engines on mobile devices usually give different results
adapted to the mobile experience. So, if we want to promote our content for
mobile users, we will have to establish a mobile SEO strategy.
Page 30
BIG DATA MARKETING
Mobile marketing has become very popular, as consumers transmit a large amount of
data through their mobile devices so the information that can be collected will very
helpful.
The application of big data to mobile marketing allows us to use and send
communication not only with the most appropriate content, but also at the most
appropriate time for each customer. In addition, and thanks to geolocation, we can
perform real-time marketing, close to the key place and for the right target.
The main mobile marketing campaigns in big data are: apps, mobile phone
communication and personalised messages.
Below is an example of mobile marketing by Starbucks and Spotify. The campaign gives
access to music lists exclusively created for customers who registered on the Starbucks
application.
Page 31
BIG DATA MARKETING
▪ Social marketing
The last type of big data marketing that will be discussed is known as social marketing.
It is defined as the technique used by social media and influencers to work on the
marketing and business needs of a company.
The type of actions that can be carried out in this marketing modality are traditional
actions such as sponsorships and social ads, or more innovative actions related to social
media such as influence marketing, and actions on the social media and blogs of the
company.
It can be said that social media is part of big data. As discussed in other subjects, social
media provide us with a huge amount of data and unstructured information about our
customers that will be really useful for the design of our marketing strategies.
To better understand the type of information that social media can give us, we will
introduce the most popular social media briefly in order to see how they should be used,
both to gather information and to carry out marketing actions.
Before going into detail, we should bear in mind that the content published by users on
social media is really important, as it gives us a lot of information about them. However,
to collect more information, the visits to that content must also be analysed, as well as
the likes and comments they have, if the content has been shared and the number of
followers or friends that our customers have on their media.
Page 32
BIG DATA MARKETING
BLOGS
Blogs are that space in which a company or individual shares their texts or articles on a
regular basis.
It should be mentioned that some people think blogs are unnecessary for a company
due to the great impact that social media have. However, depending on the sector, a
blog can be a very good resource because it allows us to show users a much more
detailed content that often complements the content shared on media. For example,
regardless of the type of company or organisation, social media such as Facebook or
Twitter can be used to present our news and include the link to our website or blog so
that users can get more detailed information of what is shown on social media. For
example, if we have interviewed someone influential in our sector, we could use
Instagram to share a photo of the interviewee and link the content with the full
interview on the blog.
The main characteristics of blogs are:
- It is a dynamic space that allows the author or authors of the content to interact
with users and readers through comments.
- Unlike a website, blogs have much more personal and informal communication.
This means that they are much closer to users, and direct and informal
communication can be established. However, we should always be respectful.
- The contents of the blog are published through posts, which are classified by
date, category and labels so that they can be identified quickly.
- Blogs can be read through RSS aggregators. This allows spreading the new
content that has been published and notifying readers.
Page 33
BIG DATA MARKETING
SOCIAL MEDIA
Social media are a social structure that consists of a set of actors (individuals or
organisations) that are related based on some criteria (professional relationship,
friendship, kinship, etc.).
Social media allow individuals to communicate and share information. They can be
divided into three levels, depending on the type of relationship of the individuals.
▪ Generic social media. The most common. Its main function is to establish
relationships among its members, which may or may not know each other
previously. The most common generic social media are Facebook, Instagram or
Twitter. They are a source of both personal and public information.
▪ Professional social media. Its members are professionally related, and these
media are used to contact colleagues at work or to seek employment. Some of
the best known are Xing or LinkedIn.
▪ Vertical or thematic social media. They are based on a specific topic and
establish relationships between users interested in that topic. Some examples
could be Flickr, where users exchange professional photographs or Dribble, a
social network for graphic designers where they show their work.
Page 34
BIG DATA MARKETING
The most popular social media that can provide us with a large amount of information
about our customers are:
Facebook is the most popular social network. All types of content can be shared on it:
short texts, long texts, images, videos and links to content from other networks or
platforms. It also allows companies to organise contests. This can bring us new
customers and can provide us with relevant information about these customers.
Two types of profiles can be seen in this network: personal accounts and professional
pages. The difference in the two options lies in the fact that in personal pages we must
add friends as users, while in professional pages users should "like" the page and follow
us to see the publications. Companies must have a professional page to gain followers
and keep them active to gather as much data and information as possible.
At the data level, we must pay attention to the comments of users on our publications,
the number of likes and dislikes and if our publications are shared by our followers. We
must also pay attention to the content that users share about us as it allows us to know
their opinions and what they expect from us.
It is a very relevant channel since it has a large amount of traffic. Through the hashtags
that are related to our company, we can access all the information and comments about
us. This will allow us to know the opinions of users about us, as well as what they want
or expect from us.
Page 35
BIG DATA MARKETING
This is the quintessential network to share and publish visual content. Photos and videos
can be shared, but they can only be published for 24 hours, and live videos can also be
made.
It is a very dynamic social network and, as Twitter, it generates a lot of data. In addition,
thanks to hashtags, it allows us to see the publications related to our company.
There is a phenomenon in this social network that must be mentioned: the so-called
influencers. They are users with a large number of followers who advertise products of
different brands on their networks. They generally talk about the product in exchange
for having it for free, but some of them charge for advertising.
Collaborating with these influencers can help us gain followers and customers. However,
before working with them, their profiles must be carefully analysed, since some of these
collaborations can have a negative result and cause the opposite effect to the desired
one.
LinkedIn is the social network focused on the professional field and on business. It allows
us to look for business opportunities, foster strategic alliances, build a good contact
network, interact with influential people, strengthen the brand, communicate the news
of the company and exchange information with users.
LinkedIn is a much more effective tool when it comes to generating potential customers,
which makes it especially interesting. In addition, the average age is much higher than
in any other social network, which generates a different audience that may be more
attractive according to our interests.
Page 36
BIG DATA MARKETING
Google +
This network is not widely used, but it should also be mentioned, since it belongs to
Google, which gives it some very important advantages. Being on it guarantees a much
more effective positioning and allows us to contact the more than one billion people
who search on Google every day. In fact, Google makes the rules in this field and their
algorithms clearly favour their own social network.
Given these five networks, it is important to discuss some aspects that will allow us to
collect data and information about our customers to help us plan our actions and
promotions.
- Geolocation must also be considered. As already said, this service will help us
know where our customers are located, and this will also help us offer them a
much more personalised service.
- We should not only consider the comments that users leave on our publications
or those of our publications that are shared by them, but it will also be helpful
to know what users publish about other brands, what they share, their tastes
Page 37
BIG DATA MARKETING
and interests, etc. This will allow us to have a comprehensive view of what they
like not only about us, but in the market.
- Content based on the study of the reaction of users to our posts. In this way, the
interests of the users can be identified and the content of the brand can be suited
to their tastes.
- Advertising according to the user's profile. This allows a better targeting of the
ads and, consequently, a higher impact.
- Posts published based on algorithms, and not directed. In this case, several posts
are written and are included in software that will publish the different articles
according to the behaviour of the users on social media or on the blog. This type
of publication has 90% higher reach and 25% more engagement than predefined
posts.
Two other Facebook and Twitter tools should be mentioned: Facebook Custom
Audience and Twitter Customer Audiences. These tools use data to carry out targeted
marketing actions. They allow companies to define different messages and actions
according to the user based on the criteria selected by the user.
These two tools have great potential, since they allow us to use the existing data of the
customer database to design ads for social media.
Page 38
BIG DATA MARKETING
However, not all actions can be automated since there are commercial actions that
cannot be predicted or that are not determined by predefined actions.
- Customise the entry page of our e-commerce based on the navigation and
purchase of users, so that each customer sees a page based on their interests,
preferences and usual purchases.
In short, all the communications that are determined by defined parameters can be
automated, especially in digital channels in which the content can be automatically
varied, such as email, messages, notifications in applications, purchasing
recommendations and the personalisation of the home screen of each customer or user.
Page 39
BIG DATA MARKETING
- Optimisation: makes the most of human resources since a single person can
contact a large number of potential and current customers.
- Save time: means being more productive in routine and repetitive actions. This
allows us to spend more time on other tasks.
- CRM: allows us to have a database with all the information about our customers,
both current and potential.
Here are the steps that should be followed for a good implementation of marketing
automation.
Page 40
BIG DATA MARKETING
- The first thing that must be done before automating our actions is to define the
strategy, that is, the objectives and the actions to be automated must be
established.
- Next, the right tool must be found. For this, the following elements should be
taken into account:
o Features à must have the necessary features. We must also make sure
that we are not paying for those features that we do not use.
o Price and price system à Lastly, the price of the tool and the type of
contract will be considered.
- Once the tool is selected, it can be used. It is advisable to start with easy actions
and tasks to learn how the chosen tool works and, once we understand how it
works, we can start with more complex actions.
We will not go in depth with automation tools, but some of them will be mentioned
together with their main features to have a general overview of them.
First of all, it is important to mention that these tools are updated quickly and
frequently. These tools are:
Hubspot: is one of the pioneering tools for marketing automation. Its main feature is
that it provides a suite of tools connected to each other that helps to attract visits from
potential clients, manage the commercial relationship, increase conversions and track
sales actions in real time.
Page 41
BIG DATA MARKETING
There are some other automation tools, such as Padot, Marketo and Eloqua.
Lastly, we will deal with the steps that should be followed to implement an automation
process in our company. These steps are:
o Etc.
Page 42
BIG DATA MARKETING
o Enter email
o Register
o Etc.
If these three steps are followed, the different processes and interactions with our
customers will be clear. This will allow us to automate some tasks and actions. For
example, based on the purchase data of the customer, a thank-you email with a discount
code can be automatically sent one year after registering and having purchased at least
twice.
Page 43
BIG DATA MARKETING
After discussing some aspects related to marketing and how big data can help us carry
out personalised actions and promotions to achieve strong and lasting relationships with
our customers, this second chapter will talk about web analytics, which will help us to
better understand the behaviour of our customers to offer them better service.
Web analytics are a fundamental element to carrying out a good marketing action
through big data. It allows us to gather all kinds of information related to the behaviours
of users on a website or app, so that our strategy can be designed based on these
behavioural parameters to obtain better results.
Thanks to web analytics, we can carry out the following tasks that will help us to adapt
effectively to our customers:
Page 44
BIG DATA MARKETING
- Comparison with other websites: web analytics is not only useful internally, but
also externally, since it allows us to compare our website with other websites
related to our sector to know the situation of our competition, the type of
strategies they are using and their relationship with customers. This will help us
to redefine our strategies if needed.
Web analytics not only measures the visits on our website and its pages, but also gives
us qualitative information to know the behaviour of users. This will help us to adapt to
customers.
We should bear in mind that the main objective of big data marketing is the
personalisation of our actions to achieve a good and lasting relationship with our
customers, so qualitative information is really important.
Given these considerations, the different tools for web analytics will be discussed below.
- Census tools.
- Hybrid tools.
Page 45
BIG DATA MARKETING
As this title suggests, this first group of tools will help us measure the impact of our
online advertising campaigns. Google AdWords is the main tool responsible for
analysing advertising campaigns, but it is also worth mentioning Business Manager, a
Facebook tool.
Thanks to these tools, we can see the operation of the campaign, and we can also obtain
external information. Thus, we can have basic information such as how many people
saw or clicked on the ad, and also qualitative information, such as the profile of users
who saw and clicked on our ad or the particular elements of each platform.
Web analytics solutions based on panellists gather information from selected users. To
do this, users must have installed a programme on their device, either on the computer,
laptop, phone or tablet, which will monitor all their navigation.
Although it may seem like a very good tool, it should be mentioned that the greatest
danger of this type of tool is the possible bias in data capture. For this reason, companies
that offer this type of tool, such as Netquest with its Netrics tool, have as a major asset
the representativeness of its panellists.
The larger the panel in volume and distribution regarding sociodemographic aspects
such as age, gender or average income, among other variables, the more representative
it will be with respect to the population and, consequently, more reliable in the
information gathered.
With this type of tools, we can also obtain the same information about our site as from
the site of our competitors, so it is a perfect tool for carrying out studies and market
comparisons, or to look for potential agreements with related pages.
Lastly, the analysis of calculations of audience overlaps is another type of analysis that
can be carried out with the tools based on panellists. These calculate the common
Page 46
BIG DATA MARKETING
audience that different websites or applications have. The results of these analyses will
allow us to measure ourselves in comparison to our competitors and obtain qualitative
information from the market.
However, the depth of user behaviour in the collected digital asset is superficial,
compared to the depth that can be obtained with census tools such as Google Analytics.
Therefore, these tools will not be adequate to perform detailed analyses on purchasing
processes or the traceability of our campaigns. In addition, the available metrics such as
number of users, page views and average times, are extrapolated statistically from the
panellist and not about the reality of the users' access to these pages.
▪ Census tools
Census tools consist of marking all the visitors in our digital asset and measuring any
activity of these visitors within it. These tools include a few lines of code in each of our
pages to capture and track the user through the generated cookies.
Unlike tools with panellists, with the installation of a programme by the user, by
including the code directly on our pages, we will obtain higher quality information on
user behaviour. In addition, all users who access our website or application can be
measured, so metrics are not subject to statistical calculations, but to the actual activity.
The most remarkable advantage of census tools is the quality of data about users that
can be captured. This allows us to perform a deeper analysis in each area. We will have
improved visibility on the audience of our website and its traffic sources, the behaviour
within each page or, in the case of an e-commerce, all the information related to the
sale of our products or services.
However, we cannot know our competitors with these tools, and this makes us unable
to make a comparative analysis. That is why these tools should be combined with
solutions of panellists or other types of studies and analyses.
Page 47
BIG DATA MARKETING
For a perfect analysis, it is very important to include the code on all the pages that make
up our website. That is, if there are pages that we think are not important on our site,
such as the legal notice page, the code should also be included on them since not
including those pages in the analysis means that we may be ignoring leakage points or
they may create problems when identifying the flow and navigation of a user on our
site.
The same thing happens with the contact page, since not including it in the analysis may
mean not knowing data such as the percentage of users who enter our website to ask
or complain, and we will not know the reason why users visit our website.
Census solutions are the most common web analytics tools used by analysts because
they allow us to know more about the user, as well as to identify the problems of the
website or application.
▪ Hybrid tools
Hybrid tools offer information obtained mainly from a database of panellist and adjust
it with data collected by census tools.
Having seen tools, metrics, dimensions and segmentations will be discussed below.
Once one of these tools has been used, the value relationships that can be extracted
must be established. Metrics, dimensions and segmentations will be used for that.
Page 48
BIG DATA MARKETING
We will focus on Google Analytics, which is the most popular and the most widely used
tool in the field of analytics.
The link to Google Analytics website can be found in the bibliography so that examples
and application cases can be seen.
▪ Metrics
Some of the metrics that Google Analytics uses are page views, pages per visit and
duration.
Because the dimension has one or more characteristics, more than one metric can be
applied at a time. Following with the previous example of the city dimension, we could
also consider metrics as sessions, new sessions, new users, average session length or
pages per session, among others.
Google Analytics divides metrics into three large groups or categories. These are:
- Behaviour: deals with how these users interact within the website or application.
Page 49
BIG DATA MARKETING
- Visits: when we talk about a visit or session, we refer to the element that
encompasses the actions of the user in our website or application. Here, we will
not only count the number of users that enter our site, but it will be important
to contextualise and analyse the quality of these visits.
- Unique visitors: this metric refers to the people who visit our site. A unique
visitor is a person who accesses our website from their personal computer at 10
o'clock in the morning and returns at 5 o'clock in the afternoon on the same
computer. This person visits twice, but it will be considered a single visitor.
However, if the device is different and, instead of visiting our website from their
personal computer, they do it from the computer they have at work, it would be
considered as two unique visitors. This is because the Google Analytics
monitoring is done through cookies.
- Page views: refers to each page loaded by the visitor on our site, whether they
are different or not. We must take into account that each user will generate at
least one visit and one page view, but they can generate more visits and page
views. Thus, we can count more visits and page views than users, but there can
never be more unique users than visits or page views since this would indicate a
wrong configuration of Google Analytics.
- Unique page views: are the different pages visited by a user during a visit. Thus,
if a user accesses the page of one of our products and then accesses another
page of another product and then goes back to the first page he or she had
accessed, the number of unique page views will be two, whereas page views will
be three. Thanks to this metric, the most visited and the least visited pages can
be found. This allows us to establish the reasons for this and increase visits on
the least visited pages. If the purpose of our website is selling advertising space,
this metric will allow us to analyse the impact of advertising among visitors. For
this, page views will be divided between total visits. The indicator or KPI obtained
Page 50
BIG DATA MARKETING
with this division will indicate the depth of the visit, that is, it will show the
interest of a visitor in our website and in relation to specific pages.
- Visit duration: this metric allows us to know the time the user spends on our
website. Google Analytics calculates the time on a page by deducting two points:
initial page and next page, so that if a user only visits a URL, Google Analytics will
be unable to calculate that visit.
- Bounce rate: a bounced visit loads a specific page, but no other action is taken,
such as displaying other pages. The bounce rate generally indicates that the user
is not interested in the content of the page or that it is confusing. It is a really
useful metric that should be consulted whenever we want to optimise a
campaign landing page. This bounce rate does not necessarily have a negative
meaning in pages with content in articles or posts, since in these cases, the
analysis must be completed by measuring the scroll of users on the page, since
this will give us information about their behaviour.
- New visitor: when analysing the visitor metric, it is important to observe how
many visitors visit us for the first time. Google Analytics can make this distinction
because it marks users with different cookies, each of them with one-year
duration. In this way, it can distinguish between visitor, visit, new visit or
recurring visit, among other variables.
- Returning visitor: The relationship and the analysis between both metrics will
provide us with value regarding the strengths or strategies of the site. A high
percentage of returning visitors with respect to new visitors, will show good user
loyalty. Thus, it would be interesting to consider our current strategy in SEO or
consider a paid traffic campaign. On the contrary, a low percentage of returning
visitors will indicate that, despite attracting a lot of traffic with our campaigns or
our SEO strategies, we are not able to retain those visits.
Page 51
BIG DATA MARKETING
To conclude on metrics, it is worth mentioning that our analysis should not only focus
on the study of a metric, but on several, as this will give us a much more comprehensive
view of our website. Thus, if visits to the website are analysed, not only new visitors will
have to be considered, but this metric will have to be compared with the metric related
to returning visitors. This will indicate if there is a high percentage of loyal customers.
▪ Dimensions
These dimensions are expressed by values. Therefore, the value of gender can be
feminine or masculine, etc.
As with metrics, it is interesting to cross more than one dimension, since this will allow
us to break the information down in a more detailed way. For example, in our analysis
of new visitors, it may be interesting to know the city from which they come and also
the age of these new visitors. These two dimensions will be crossed to see the different
age ranges of our new visitors, so that this will allow us to know what audience to target
and what strategies to use for it.
Google Analytics establishes a wide range of dimensions for analysis. However, this tool
can work with customised dimensions so that this allows us to better adapt to the needs
of our company. For example, if we have an online store, we can distinguish between
the users who register and then purchase, the users who register at the time of
purchasing and the users who buy without registering.
In addition to knowing the percentage of users of each type, these dimensions allow us
to know aspects such as the number of registered customers who buy, so we can relate
this to the brand's loyalty.
Page 52
BIG DATA MARKETING
values and that can have results that can help us to know the real situation of our
company.
▪ Segmentations
Segments can simply be the limits of a dimension, the combination of values between
different dimensions or the combination with different metrics.
Thus, we can create a segment among all the users that connect through a specific
device, and distinguish between users that connect through a mobile phone, tablet or
computer. We could also segment based on other variables, such as age or city of
residence.
Thanks to segments, subsets of data can be isolated and segmented, and trends within
our site can be extracted.
In the acquisition section, Google Analytics lists all the traffic sources and groups them
in different dimensions, such as source, medium and campaign. In this way, it allows us
to compare among the results of different campaigns as thoroughly as we want to check
the performance of each channel. To do this, all our campaigns must be parameterised
correctly, so that they can be differentiated in Google Analytics.
Page 53
BIG DATA MARKETING
Google Analytics shows us by default a group of traffic channels based on rules of the
most common traffic sources. These are:
CHANNEL DESCRIPTION
Organic search Traffic coming from search engines, from organic results
(free).
Social network Groups sources considered social media that are not
labelled as ads.
Referral Traffic coming from websites that are not social media.
They can be blogs, forums and similar.
Direct It shows the traffic of users who wrote the URL of the
website in the browser or who came to the site through
a bookmark, such as favourites.
Page 54
BIG DATA MARKETING
In addition to knowing the channel through which users arrive, we will probably want to
analyse the traffic coming from these customer acquisition channels in more detail. We
are not only interested in knowing where they come from, but also in more specific data,
such as what information in the newsletter resulted in more traffic to our website or
better sales performance.
To obtain this type of information, we must parameterise the URLs of the links of each
campaign adding specific parameters for Google Analytics called UTM.
When analyzing the results obtained from web analytics, the following four types of
analysis must be considered:
- Audience
- Acquisition
- Behaviours
▪ Audience
As its name suggests, this first analysis gives us information about the audience, that is,
about the visitors of our website. The information obtained in this analysis allows us to
better know our users and it also helps us to obtain useful information for the
segmentation of our online campaigns.
The types of reports that can be obtained with audience analysis are:
- Demographic data: refers to the age and sex of our visitors so that we know the
percentage of men and women interested in our site, as well as the percentages
in the different age ranges. One of the main advantages of having this
Page 55
BIG DATA MARKETING
information about our audience is that the content of our page can be
personalised. If our website contains advertising, we can also segment ads,
which will increase the relevance for our audience.
- Technology: One of the key elements for the development of our website or
application is the optimisation by device and its testing. Users can connect and
access our website from different devices with different sizes, so our website or
application must be adapted to each of them. Currently, optimisation by device
has become less important because many websites or applications use
responsive designs, that is, designs that adapt to all types of screen and device
dimensions. However, there are some websites or applications that do not use
this design, so it is important to know which devices, browsers and versions are
most used by our visitors, as this will help us to plan development. The speed
and the type of connection should also be considered, since multimedia contents
will have to be adapted based on them. A key aspect to consider in this report is
the quality assurance process, which refers to the process of testing that
everything works properly. Variables such as the size of the screen, the browser,
the connection and the operating system will be taken into account for this.
Page 56
BIG DATA MARKETING
Thanks to the technology report, we will be able to identify the most common
configurations among our users and ensure that these configurations meet our
quality standards. In short, this is a truly important analysis to adapt our website,
as well as to improve the user experience to improve retention or conversion of
users.
- Group analysis: this report gives us a cohort vision, that is, a vision of a group of
people who share a characteristic or experience within a certain period of time.
This will allow us to know if this group adds value to our business in the long
term. An example of this analysis could be to analyse all users with the same
acquisition date, for example, during January, and see how their purchases
evolve in the following periods. Another example would be to analyse customers
who purchased their products during sales period, monitoring them during the
following six months in order to establish if they are customers who only buy
because of sales or if instead, they also buy when there are no discounts. In these
analyses, it is important not only to see if the purchase action is repeated, but
also the time it takes customers to buy again, the times they buy, if they buy the
same type of product or if it is a totally different product, or the average amount
of the purchase, among other factors. We should use these reports to compare
the different groups.
Some other interesting reports regarding the audience are the report of active users,
and the report related to user loyalty in terms of whether they are new or returning
visitors, as well as frequency and hook reports. Customised variables about our users
can also be collected, so that reports of more specific aspects can be made. As usual,
before customising the variables, it is very important to consider if the variable we want
to analyse can provide us with useful and valuable information that will help us improve.
Page 57
BIG DATA MARKETING
▪ Acquisition
The analysis of acquisition will help us know the origin of our users according to the
channels used and the campaigns carried out. It is a really valuable information when
analysing and optimising our online marketing campaigns.
- Overview: with Google Analytics, this first report gives us an initial picture of the
channels that drive more traffic to our website or application. From this general
report, some more specific reports about the origin of our users can be
extracted.
- Source and medium: This report gives a list of all traffic sources, regardless of
the channel they belong to, in addition to giving us more specific information
about the behaviour and conversions of these sources than in the overview
report.
- Campaign: The campaign report gives more detail since it gives the same
information but divided by campaign. For more detail, the channels report can
be consulted. In any case, regardless of the report, information can be added
and broken down in any of them, which adds a secondary dimension to the
generated report.
Page 58
BIG DATA MARKETING
- Social media: as its name suggests, this report will give us information about the
social activity related to our website or application. We will not only get a
breakdown of traffic or conversions from different social media, but we will also
have information on the engagement of users by measuring shares or likes.
- Search console: these reports analyse the performance of organic search traffic.
They provide us with related information about the number of times our site and
our pages appear in search results, as well as the bounce rate of SEO landing
pages. The search strings of users can also be seen, that is, the keywords they
used in search engines and the average position in the Google results page for
those words.
▪ Behaviour
This type of analysis gives information about what our visits do when they land on our
website or application and what content is the most successful.
In this type of report, the pages of our site or the screens of our application are the point
of reference for information. The metrics of page views and unique page views will be
seen there, and they also include the time the user spends in each one.
- All pages, landing pages and exit pages: this first report shows all the pages that
our website contains along with values such as the number of page views, the
average time on each page, the entry pages, the bounce rate and the percentage
of outputs of each of them. Landing pages are the pages where visitors first land,
while exit pages refer to the last page visited by the user before leaving.
- Events: Google Analytics events give very important information about the
behaviour and interactions of users on our website. Actions such as clicks on
certain buttons can be seen, as well as additional information on interaction with
forms or advanced behaviours. Events are a very good option to monitor specific
actions carried out by the users. They also allow us to record percentages of
Page 59
BIG DATA MARKETING
video viewing or page reading. This shows the time a user spent on a video as
well as the exact amount of content that was read thanks to the scroll control,
which allows us to identify the content that users have not seen. Events also
focus on the forms of our website, so that we can know what users have filled
out in a form before leaving or which fields cause more problems, so that these
forms can be improved to be entirely filled out by users.
- Site searches: this is a report on the internal search engine of our website. Some
of the information that can be extracted is: the most searched keywords, the
number of users who have made a search, the pages from which the search was
made or the landing pages after the search, among others. In this report, the
search information on the site can be very valuable when it comes to finding out
what our users really want to find once they have reached our page. Thus, we
will identify the most popular products or the products that should be
highlighted in our results grid, because most users are looking for them. We will
also find out the products that users look for on our website and cannot find, so
this will help us when we want to add more products.
Through the objectives, we can measure how, in our web or application, certain actions
occur or an activity called conversion is completed, which will be defined based on our
interests. In this way, our site and our marketing campaigns will be assessed based on
the business objectives we set.
Purchases are usually the main objective of the sites or applications that generate
income. However, other objectives that should be highlighted are the completion of
registration forms, the download of product catalogues, clicks on the buttons that show
contact data, time on the site and see and read a number of items or posts, etc. They
will depend on the type of company.
Page 60
BIG DATA MARKETING
This section will focus on analytics for e-commerce. For this, the following four points
will be discussed:
- Conversion funnels
- Improved e-commerce
- Attribution models
Each of these points will be discussed below in order to know how to carry out analytics
in e-commerce.
Continuing our focus on Google Analytics, we will deal with the basic analytical metrics
and reports that will allow us to obtain relevant information about our online commerce
in order to help us in the design and implementation of strategies and marketing
campaigns.
Page 61
BIG DATA MARKETING
website or application involves. There will be all the information regarding transactions,
as well as the income breakdown by type, the products purchased and the days since
the user accessed the site and purchased, among other elements.
The main metrics and reports about our e-commerce that can be obtained are the
following:
- Overview: the overview report is a global summary of the revenues of our online
commerce, as well as of the number of transactions completed and their average
value. Google Analytics considers that a completed transaction is a purchase
order that contains one or more products, regardless of who performs this
process. So, if the same user makes two purchases, they will be counted as two
transactions, even if it is the same buyer. To calculate the average value of the
transactions, the expenses corresponding to the transaction must be added to
the price of the product or service that the customer buys, such as the shipping
costs. Finally, the percentage of conversion of electronic commerce can be
obtained in this first report; in other words, data referring to the number of
transactions made and the number of sessions so we will know how many of the
users who accessed our site bought and how many did not.
- Sales performance: in this third report, transactions will be the reference point.
They will be broken down into the following items: order entry, taxes and
Page 62
BIG DATA MARKETING
shipping costs associated with each purchase, as well as the quantity of products
of each order.
In addition to these three reports, reports on other relevant data can be generated, such
as the time it takes until the user buys, that is, the number of days between users
accessing our site for the first time until they place an order, and also the number of
visits they needed before deciding. To complete this report, it is interesting to compare
this information with the customer journey. As we already know, the customer journey
is a document that shows the customer journey during a specific process or interaction
which allows us to know their behaviour and identify the moments that can create
doubts or insecurities to make the necessary changes and improvements so that the
purchase process is faster.
Conversion funnels allow us to see the steps that users take to complete a task so that
the points that generate a higher abandonment rate can be identified.
In this way, the customer journeys that are ineffective or have been abandoned can be
reduced, and opportunities can be identified. For example, if many users leave in a step
of a purchase process and go to the page of frequently asked questions, this means that
there is something in that step that is confusing to users. It will be necessary to identify
what creates doubts and solve it so that when customers are at that step they do not
need help.
The most frequent use of conversion funnels is the analysis of the steps of a purchase
process. On the basis that we want a product to be purchased, we must design the
Page 63
BIG DATA MARKETING
funnel including all the steps to complete the purchase process on our website, that is:
shopping cart, personal data, payment method, data and shipping method, etc.
We must bear in mind that the funnel should be as narrow as possible, so it should
include all the steps of the process in a detailed and specific way, since this will give us
more information that can help us draw better conclusions. We should ideally include
all the processes, both the main and the secondary processes to identify problems in
any of them.
Once we have designed the funnel and we have the results, we can use two tools to
analyse the funnel. Although they are very similar conceptually, they have significant
differences that must be known to interpret the data in the best possible way. These
tools are:
- Goal flow
We will discuss how to use each tool, as well as its main features.
The first thing to note about these two tools is that, in the conversion funnel chart, a
session is shown in each step of the conversion funnel. This means that if a user goes
through the same step twice, either going forward or backwards to make sure of a point
or to refresh a page, this second session is counted as an exit from the page of that step.
On the contrary, with the goal flow tool, the steps backwards of the user can be seen
so that it helps us identify the weak points in any funnel.
Advanced segments cannot be used with the conversion funnel chart, while they can be
used with the goal flow. The application of the segments allows us to know the exact
behaviour of very different profiles in a funnel.
Page 64
BIG DATA MARKETING
For example, if the goal flow is segmented on new or returning users, and we make any
change in the step process, we can see if this change affects both types of users equally.
If a user, regardless of the reason, skips a step between the access step to the funnel
and the exit step, the conversion funnel chart tool will fill in the steps that the user
skipped, that is, steps that the user has not visualised will be drawn. If a user goes from
step 1 to 3 in a process of three steps skipping the second step, the conversion funnel
chart tool will reflect that the user went through the three steps. On the contrary, if we
use the goal flow, the steps that the user skips are not reflected.
Page 65
BIG DATA MARKETING
chart-ba21bcba-f325-4d9f-93df-97074589a70e
There is another difference between these two tools regarding the definition and user
journey within a funnel. While the real order in which the steps of the process have been
seen is not reflected in the report of the graphical funnel tool, it is reflected in the goal
flow.
A wide range of metrics and dimensions that inform us about the situation of our online
commerce have been discussed so far.
However, Google Analytics not only provides us with key information about conversions,
the number of products sold, revenues, etc., but it also provides us with other types of
reports related to the buying behaviour of our visitors. Thus, the data obtained through
these reports will affect our catalogue.
- Buy-to-detail rate: How many products were purchased per number of product-
detail views. These data will show us which products are the most likely to be
purchased once the product details page has been seen. This metric will be more
or less clear and will have more or less relevance since this will depend on the
type of product. Elasticity will be lower for low value, standard or little
differentiated products, since users do not need to consult their details, while it
will be higher for more complex and valuable products. If we want to obtain a
Page 66
BIG DATA MARKETING
high buy-to-detail rate, the design of our e-commerce is very important since the
way it is designed can promote and influence users to visit the product page or
show it on the results grid.
- Product adds to cart and product removes from cart: As the name of this report
suggests, the data that will be obtained here refer to the times a product is added
to and removed from the purchase. This allows us to know which of our products
or services have a higher conversion rate, and those which have a lower
conversion rate. However, we should not only focus on the data obtained, but
we should also analyse the reasons that can make users remove a product from
the cart. For example, if some of our products have a low price, but customers
must pay a shipping cost of 5 or 6 euros, it may not be worth buying it. Another
reason may be the time it takes for the product to arrive, which may be
considered too long by the customer. This report should not only show us those
products that are added to or removed from the cart, but we should try to assess
all the factors that make the customer decide not to buy the product.
In addition to these three main reports, other reports will be very useful, such as internal
promotion clicks and views, which allow us to monitor our internal promotions and know
the most visited and the most clicked promotions.
The product refunds report is also very interesting. It is related to other departments of
our company such as the department of quality and customer service and will show us
the products that have a higher rate of return.
Page 67
BIG DATA MARKETING
The order, product or affiliate coupon codes are some other reports. The latter allows us
to see affiliate websites that contribute to sales.
To conclude this subject, attribution models will be discussed. Attribution models are
the rules that determine how credit for sales and conversions is assigned to touchpoints
in conversion paths.
For example, the last interaction model assigns 100% of the value of sales and
conversions to the final touchpoints (i.e., clicks) that immediately precede sales or
conversions. On the contrary, the first interaction model assigns 100% credit to
touchpoints that initiate conversion paths.
Google Analytics offers some predetermined attribution models. The most common way
to work is just with one model, but it is interesting and also necessary to compare the
results with several attribution models, as this will help us not to ignore traffic channels
that may not seem useful, but that clearly contribute to conversions.
Page 68
BIG DATA MARKETING
MODELS DESCRIPTION
Last interaction The last touchpoint receives 100% of the credit for
the sale.
Last Non-Direct Click All direct traffic is ignored, and 100% of the credit
for the sale goes to the last channel that the
customer clicked through from before converting
Last Google Ads Click The last Google Ads click receives 100% of the credit
for the sale.
First interaction The first touchpoint receives 100% of the credit for
the sale.
Position based 40% credit is assigned to each the first and last
interaction, and the remaining 20% credit is
distributed evenly to the middle interactions.
Page 69
BIG DATA MARKETING
3. LOYALTY PROGRAMMES
This third chapter will talk about the loyalty programmes carried out by companies to
achieve customer loyalty.
Big data marketing uses a wide range of tools and parameters to get to know customers
perfectly so that they can be offered customised products and promotions based on
aspects such as their tastes, needs, expectations and behaviours.
Carrying out personalised strategies will help us to increase the level of satisfaction of
our customers as well as to retain them. This will mean that they will come to us and
not to the competition when they want to purchase a product or service.
Marketing first considered that the best way to achieve customer loyalty was with new
products, different from the products offered by competitors. This stage was really
important, and actions such as copy, imitation and improvement of the products stood
out. In other words, companies tried to improve the products of their competitors to
offer different products to consumers and win their loyalty. In fact, companies still work
like that, but a period of knowledge and understanding of the customer is added to this
process to create and develop new products. Loyalty programmes are created in those
periods to consolidate the market share of the company.
Page 70
BIG DATA MARKETING
Customer loyalty is not just about offering a product that is better than the competition,
nor offering a product adapted to each segment. Customer loyalty is achieved by
building a relationship between customers and the company through which value is
added to customers and they see it as such.
When creating a loyalty strategy, two basic elements should be taken into account.
These elements are:
- The acquisition a new customer is usually five times more expensive than
retaining a current customer, which involves a high investment in marketing,
many efforts of different teams of the company and the uncertainty of not
knowing what type of customer is being acquired.
Page 71
BIG DATA MARKETING
- The profitability of a customer increases as its life with the brand increases since
reducing customer attrition by 5% can increase the profit between 25 and 125%.
- The reasons for customer abandonment or customer attrition are not usually
related to the product or service itself, but to the lack of engagement and the
quality of the service.
After having seen the importance of customer loyalty for our company, the different
loyalty programmes that can be used will be explained below.
Currently, there are different types of loyalty programmes with different value
propositions and rewards policies.
Each company uses different supports for these programmes, such as clubs, applications
or membership cards, which will enable customers to obtain rewards. These rewards
can be of two types: rational, such as discounts, special offers and exclusive prices for
members, or emotional, such as specific benefits for members, personalised
communications or free courses and activities for members.
Page 72
BIG DATA MARKETING
- Multi-partner programmes
▪ Multi-partner programmes
This type of programme arises when a group of companies create a new brand/company
that deals with defining and managing the entire club. This means that the main
advantage of this loyalty programme is that the customer can earn points for rewards
when buying in different associated companies.
This type of programme usually works as follows: points are earned for the purchase of
a product or service of the different associated companies, and these points can be later
exchanged for awards, such as products of the same companies, discount vouchers, etc.
In addition, they are based on the fact that the associated or sponsor companies
perform three main tasks. These are:
- Sponsor companies pay for the points they send to their customers.
However, these programmes also have disadvantages such as the fact that the individual
company loses its strength, the calculation of rewards is very complex and great effort
is required to obtain aspirational awards such as trips or leisure offers.
Page 73
BIG DATA MARKETING
Programmes based on the type of acquisition can be divided into the following
programmes:
The loyalty strategy is the foundation of any club since this strategy allows companies
to differentiate themselves from each other and establish the different rewards.
- Strategy of award points without time limit: as customers buy, they earn points,
and when a point limit is reached, they receive a reward.
- Strategy of award points with time limit: customers earn points when buying
during an established time. If they earn a specific number of points before the
time limit, they receive a reward.
Page 74
BIG DATA MARKETING
Companies offer this type of programme so that their customers can finance their
purchases and do not have to pay the full amount at the time of purchase.
The following advantages of the applications as support for loyalty programmes should
be highlighted:
- Cost reduction
Having seen the different types of loyalty programmes, the next section will deal with
the strategy of a loyalty programme. As said, it is very important since it should be in
accordance with the type of company and its philosophy.
Page 75
BIG DATA MARKETING
- Competitive position: What is the position of our company in the market? What
strengths and weaknesses does it have?
- Customer voice: Does our company listen to what customers say? What tools
are used to understand the opinions of our customers, and how are they used?
- Establish what the company is focused on: customer initiatives and customer
objectives assigned to management.
- Know and analyse what our customers say from a qualitative and quantitative
perspective.
If we follow these three steps, we will obtain the diagnosis of the initial situation on
which our strategy will be defined. Customers should be clear on the value proposition
that will be obtained if they belong to the club. The architecture of the club should
establish its management and profitability so that it must establish how to implement
actions or policies that allow giving the promised advantages and benefits to customers.
As we know, customers are one of the essential elements in our marketing plan so a
customer strategic plan will be necessary. This plan refers to planning all the actions that
the company will carry out according to customer segments or micro-segments and the
strategy defined for each customer.
The objectives established for each segment may be: retain, develop, reactivate or
increase frequency, among others. Depending on them, specific actions will be designed
Page 76
BIG DATA MARKETING
that will establish the objective to achieve, the monitoring metrics and the budget for
each action.
For example, if some customers have not bought for a while and we want to reactivate
them, we could send them a personalised email reminding them of the most purchased
products during a specific time period in the past and offering a discount coupon for
their next purchase, which will have to be made within a certain time. The metrics that
can be used are: the percentage of open emails, the number of clicks on the email link
that leads to the online store or the number of times that the promotional code has
been used during the time that the company established as a purchase period.
This strategic plan is annual and is based on the knowledge of customers. It applies both
the transactional information such as sales data and customer receipts, and qualitative
and quantitative information.
The following four steps are necessary to define the strategic customer plan:
- Resource allocation
As we know, each company consists of several departments or areas with very clear and
specific roles and tasks. This makes each of them set certain objectives to be achieved
during a certain period of time that can be a year or less than a year.
Page 77
BIG DATA MARKETING
However, although each department has its functions, it is very important that these are
coordinated so that the objectives set by each department should be consistent with
the others. For example, the sales and customer service department aims to increase
the satisfaction of its customers by highlighting the benefits and advantages of its
products, but the manufacturing department aims to work with less expensive raw
materials trying that this does not affect the quality of the product, but considering that
this may happen.
These two objectives can be opposed, since one of them is intended to increase
customer satisfaction. But, we also want to manufacture the same product with
different materials that may affect the quality of that product. It is very important that
the objectives are established in a coordinated way to ensure they are not conflicting.
The corporate strategic plan must be developed before setting the specific objectives of
each department. In this way, all departments know the status of the company in each
department as well as the global objectives, and are able to set specific objectives
according to the current situation.
This diagnosis will help us to further know our customers. A customer centric company
should always make their decisions based on customers since their opinions, needs and
comments should be considered to be able to correct, improve and increase their link
with the brand. In other words, our strategy must be based on what our customers think,
want and expect.
Page 78
BIG DATA MARKETING
Companies can use several tools to conduct a good business and customer diagnosis.
These are:
▪ Surveys: are a quick and easy way to collect data and prepare studies in order to
use the results and apply improvement processes.
Page 79
BIG DATA MARKETING
These tools allow us to analyse our customers to get to know their opinions, needs and
what they expect from us. Once this information has been analysed in detail, we can
define the strategies and actions that will be performed on the different types of
customers our company has.
In addition to the strategies, we must consider aspects such as the level of effort, the
budget and the resources that will be allocated to each group. As seen in the module
about Customer Analytics, there are more or less valuable customers, so the elements
allocated to each group will be strictly related to the value of each customer.
The customer strategic plan consists of different policies and each of these policies has
a wide range of both promotional and communication actions.
▪ Loyalty policy: aims to reward loyal customers, that is, to reward our best
customers. Rewards for previous purchases are the most frequent rewards.
Page 80
BIG DATA MARKETING
- Resource allocation
To finish the customer strategic plan, resources, including material and non-material,
will be allocated.
It is very important to have planned some actions and strategies according to the value
of each customer or group of customers, since, as previously mentioned, it is very
important to retain the most valuable customers. Companies generally allocate a higher
amount of resources to the actions aimed at the most valuable customers. The amount
of these resources usually decreases as the value of customers decreases.
When we talk about the architecture of a loyalty programme, we refer to the set of
processes and resources assigned to the management of the club that allow the give the
customer the benefits of the club easily and effectively.
The main purpose of the architecture of a loyalty programme is to ensure that customer
management is aligned with the business objectives of the organisation, taking into
account system processes and an adequate change management.
This third section will deal with the different processes and resources that make up the
architecture of a loyalty programme.
When defining the business case for a club or loyalty programme, the following
elements should be taken into account:
Page 81
BIG DATA MARKETING
o Launch campaigns
o Training sessions for the staff that works with customers, whether in
person or telematically.
- Operating costs:
The most important element that must be considered when deciding on a loyalty plan
is profitability. It will be important to calculate the ROI (return on investment). For this,
once all the previous costs are analysed, the sales that the plan will result in should be
calculated.
Based on the knowledge of our customer base, we should be able to break down
incremental sales according to the strategy for each segment:
- Acquisition
- Abandonment reduction
Once the expected net income and the total associated costs for differences are
calculated, the margin will be obtained.
The limit cost is the maximum amount that can be spent to obtain our net income
objective.
Page 82
BIG DATA MARKETING
The break-even point occurs when the benefit is equal to 0. This means that all the
margin obtained is only used for investment in order to achieve the desired objective.
To obtain the net income objective, the ROI will be calculated. It is important to mark
an objective ROI, for example 50%. This means that we want a return of half of the
investment made.
Customer segmentation involves dividing our customers into parts; each part is
characterised by some features that make them different from the rest.
We can first segment our customers based on the data we have on hand, such as
descriptive data or behavioural data.
This analysis shows to what extent this paradigm is real in our case and locates each
customer in their step of the value pyramid.
The Pareto principle is based on empirical knowledge. The Italian economist Vilfredo
Pareto concluded that people in society seemed to divide naturally into what he called
the “vital few,” or the top 20 percent in terms of money and influence, and the “trivial
many,” or the bottom 80 percent.
The RFM analysis consists of classifying customers based on their value according to
three variables:
Page 83
BIG DATA MARKETING
Internal information
The most common segmentations that can be done through internal information are
the following:
It is worth mentioning that if these segments are enriched with external information,
we will be able to increase the knowledge of the segments and extrapolate market data
such as the purchasing power index, the type of consumption, etc.
External information
As it has just been said, external data allow us to better understand our customers
thanks to the sociodemographic and behavioural contextualisation of their transactional
information.
Page 84
BIG DATA MARKETING
The sources of external information that should be considered are the following:
- Big data: refers to information related to systems that manage large amounts of
data (wireless sensors, mobile devices, histories of applications, cameras,
microphones, radio frequency readers, etc.)
Based on all the internal and external information, we can generate customer segments
focused on different aspects that describe and differentiate them. The most common
aspects for this segmentation are:
Page 85
BIG DATA MARKETING
- Demography
- Behaviour
- Transactions
- Life cycle
Segmentation will:
- Increase the effectiveness and profitability of the analysis and the development
of targeted marketing actions: adequate product and price, adequate channel,
adequate message, to the right customer.
- Identify geographic areas with higher potential market density by segment and
develop strategies for an optimal location of the points of sale.
In short, we should segment in the most specific way possible so that this allows us to
successfully reach all types of customers and achieve their satisfaction and loyalty to the
company.
Page 86
BIG DATA MARKETING
The policies of a loyalty programme were previously discussed, so this section will deal
with the elements that should be taken into account when setting up a club or loyalty
programme.
Content
Content can be of two types: rational content, which is oriented to discounts and
promotions, and emotional content, which is aimed at exceeding expectations and
generating emotions or feelings.
We should consider if the level of rewards will be low, medium or high. In addition, we
will determine the time used to obtain and use this reward, which may be immediate or
not.
3.3.4 Omnichanel
Page 87
BIG DATA MARKETING
us to improve the customer experience, in this case, the customer that is a member of
our club.
To establish the channels that will be used adequately, we first need to identify the
touchpoints that we want to use to deliver the content of the club. In other words, we
must establish the points that require customer-company interaction and through
which channel or channels we want to send the information to our customers.
The communication channels of the club or loyalty programme are all those that are
available to the company. Thus, traditional channels such as telephone or post will be
considered, but social media, email, and mobile marketing will also be included.
In the case of having a physical establishment, we must keep in mind that this
establishment is also a communication channel that often becomes the first contact of
the customer with the company.
When establishing the communication channels to send our communications and offer
promotions to our clients, we should know which channels our customers use the most,
and through which channels they prefer to receive the information. In fact, many loyalty
programmes ask customers for the channels they prefer at the time of registration, so
they are adapted perfectly to them. We will use those channels that our customers ask
us for. In other words, if a customer prefers to receive information only on their mobile
phone, we will respect that, just as if someone does not want to receive anything and
prefers to be informed in the store, since they visit often.
In addition to this, we must bear in mind that not all messages are valid for all channels,
so we must study the available channels, the messages we want to send and the type of
customer we are targeting very well. This will greatly help us when establishing the
channels to be used.
For example, young people use their smartphones very frequently so it makes sense to
use apps or email to connect with them and inform them of promotions and news.
Page 88
BIG DATA MARKETING
However, it would not make sense to do it through post since our information would
probably not reach them.
We must not forget that we try to make big data marketing actions as personalised as
possible, so we must always consider customers, as well as their opinions, tastes and
preferences, in order to achieve their complete satisfaction.
The last phase in the definition of the architecture of a club or loyalty programme is to
establish the KPIs of the customer and create a dashboard and reporting system that
allow us to monitor the variables to analyse and thus assess the success of the
programme.
The dashboard is the tool that will be used to track the variables we want to analyse.
The dashboard helps us to quickly understand the complexity of the available data since
it gives us a general idea of what we have in order to analyse more specific aspects.
- In addition, these indicators or KPI must be actionable for the business, that
is, they must allow us to carry out strategic actions to improve.
- As said, it is very important that the data and information are shared
throughout the company, so its visualisation should be as clear and easy to
Page 89
BIG DATA MARKETING
Companies can use many tools to develop a dashboard. Data should be entered, and it
will then be converted into graphic information to facilitate its interpretation. This
subject will not be discussed here since it has already been dealt with in the Customer
Analytics module and the ETL Process.
The metrics or variables to be dealt with in the dashboard of a club or loyalty programme
are the following:
- Redemption cost.
Page 90
BIG DATA MARKETING
The elements that should be considered in the design of a loyalty programme have been
discussed so far. However, there is still an element to which attention must be closely
paid. This element is the customer base.
Our customer base must be periodically analysed to see how it is evolving. In this
analysis, the following questions must be answered:
- How are sales distributed in the different segments? What sales margin does
each segment provide?
Page 91
BIG DATA MARKETING
Customer segments
As already mentioned, not every customer has the same value or the same potential, so
it is very important to identify the value of each segment to correctly assign the budget
and resources for each one. For this analysis to be successful, we must have some basic
data that allow us to identify the importance of each segment.
Through the information obtained by analysing the segments, we can adapt the offer,
the allocation of resources, define actions and campaigns, etc.
As already mentioned, we can also cross the internal data with the information coming
from external sources to complete the analysis since this will allow us to know the share
of wallet that we have in our customers regarding the market.
Customer status
A customer status is an attribute of the database that tells us in which life cycle that
customer is.
The segmentation of the customer base allows us to take a static picture of our
customers, but this indicator shows the evolution of each segment.
- New: when the customer makes their first transaction with our company.
- Active: This status is given if the customer purchases during the analysis period.
There are three sub-states in this status:
Page 92
BIG DATA MARKETING
- Sleeping: when the customer has not bought for a certain time.
- Inactive: after the sleeping status, when the customer has not bought for a long
time and is not interested in doing it again.
Acquisition channel
As mentioned, acquiring new customers is more expensive than retaining them, that is,
the customer acquisition phase is the most difficult and costly. Regarding the analysis of
our customer base, it is very important that the reason for a customer to register is
identified, as well as the channel used for it.
It is particularly important to know the customers who have signed up because they like
our company and our products, and the customers who have done it because of a
discount or promotion, as this will help us to know the most effective ways to attract
customers and the ways that add more value to the customer base.
The main reasons that lead our customers to become inactive must be identified, as well
as the effect that this causes in the different analysis indicators.
Knowing the reasons why our customers unsubscribe from the club is really important
since this will help us to improve and avoid repeating our mistakes.
This indicator measures the level of loyalty of our customers. This can be general, at the
brand level, or specific, dealing with each touchpoint, so that the weak points in our
customer management processes can be identified and the necessary changes to
improve it can be made.
This analysis must consider the customer lifetime value (CLV), which is a prediction of
the net profit attributed to the entire future relationship with a customer.
Page 93
BIG DATA MARKETING
Companies use this indicator to measure the current value of a customer's future
benefits. This indicator consists of three elements or key factors:
- Margin generated by the customer (annual benefit per customer less the annual
retention cost per customer).
- Retention rate.
As marketing has become more targeted and data oriented, a much more dynamic
approach for the analysis of the results of the CLV has emerged. This is no longer focused
on the transactional historical information of a customer, but on the future relationship
that a company can have with that customer.
As a company, the CLV calculation for each of our customers gives us the ability to:
- Identify the customers with whom we are wasting our internal resources.
Thus, using the CLV will make us better sellers because this information tells us how
much we can actually spend to acquire or retain a customer.
Page 94
BIG DATA MARKETING
4. ANNEXES
This section will deal with three real cases in which big data marketing has been used,
as well as with some examples of successful loyalty programs.
4.1 NETFLIX
Link: https://www.smartdatacollective.com/big-data-how-netflix-uses-it-drive-
business-success/
Legendary Hollywood screenwriter William Goldman said “Nobody, nobody – not now,
not ever – knows the least goddam thing about what is or isn’t going to work at the box
office.” He was speaking before the arrival of the internet and Big Data, and since then,
the streaming movie and TV service Netflix has based its business model on attempting
to prove him wrong. Netflix is said to account for one third of peak-time internet traffic
in the US. Last year it announced that it had signed up 50 million subscribers around the
world. Data from all of them is collected and monitored in an attempt to understand our
viewing habits. But its data isn’t just “big” in the literal sense. It is the combination of
this data with cutting edge analytical techniques that makes Netflix a true Big Data
company. A quick glance at its jobs page is enough to give you an idea of how seriously
data and analytics is taken. Specialists are recruited to join teams specifically skilled in
applying analytical skills to particular business areas – personalization analytics,
messaging analytics, content delivery analytics, device analytics… the list goes on.
However although Big Data is used across every aspects of the Netflix business, their
Holy Grail has always been to predict what its customers will enjoy watching. Big Data
analytics is the fuel that fires the “recommendation engines” designed to serve this
purpose.
Page 95
BIG DATA MARKETING
Efforts here began back in 2006, when the company was still primarily a DVD-mailing
business (streaming began a year later). It launched the Netflix prize, offering $1 million
to the group which could come up with the best algorithm for predicting how its
customers would rate a movie based on their previous ratings. The winning entry was
finally announced in 2009 and although the algorithms are constantly revised and added
to, the principles are still a key element of the recommendation engine. At first, analysts
were limited by the lack of information they had on their customers – only four data
points (customer ID, movie ID, rating and the date that the movie was watched) were
available for analysis. As soon as streaming became the primary deliver method, many
new data points on their customers became accessible. Data such as time of day that
movies are watched, time spent selecting movies and how often playback was stopped
(either by the user or due to network limitations) all became measurable. Effects that
this had on viewers’ enjoyment (based on ratings given to movies) could be observed,
and models built to predict the “perfect storm” situation of customers consistently
being served with movies they will enjoy. Happy customers, after all, are far more likely
to continue their subscriptions. Another central element to Netflix’s attempt to give us
films we will enjoy is tagging. It pays people to watch movies and then tag them with
elements that the movies contain. It will then suggest you watch other productions
which were tagged similarly to those which you enjoyed. This is where the sometimes
unusual (and slightly robotic-sounding) “suggestions” come from – (“In the mood for
wacky teen comedy featuring a strong female lead?”) It’s also the reason that the service
will sometimes (in fact in my experience, often!) recommend I watch films which have
been rated with only one or two stars. This may seem like it is counterintuitive to its
objective of showing me films I will enjoy. But what has happened is that the weighting
of these ratings has been outweighed by the prediction that the content of the movie
will appeal. This article explains the science behind that process in a bit more depth, and
how Netflix has effectively defined nearly 80,000 new “mirogenres” of movie based on
our viewing habits!
Page 96
BIG DATA MARKETING
More recently, Netflix has moved towards positioning itself as a content creator, not
just a distribution method for movie studios and other networks. Its strategy here has
also been firmly driven by its data – which showed that its subscribers had a voracious
appetite for content directed by David Fincher and starring Kevin Spacey. After
outbidding networks including HBO and ABC for the rights to House of Cards, it was so
confident that it fitted its predictive model for the “perfect TV show” that is bucked
convention of producing a pilot, and immediately commissioned two seasons
comprising of 26 episodes. Every aspect of the production under the control of Netflix
was informed by data – this Wired article explains how even the range of colors used on
the cover image for the series was selected to draw viewers in. The ultimate metric
which Netflix hopes to improve in the number of hours that customers spend using its
service. You don’t really need statistics to tell you that viewers who don’t spend much
time using the service are likely to feel they aren’t getting value for money from their
subscriptions, and there is a danger they will cancel their subscriptions.
Quality of experience
To this end, the way that various factors affect the “quality of experience” is closely
monitored and models are built to explore how this affects user behavior. Although its
vast database of movies and TV shows is hosted internally on its own distributed
network of servers, it is also mirrored around the world by ISPs and other hosts. As well
as improving user experience by reducing lag when streaming content around the globe,
this reduces costs for the ISPs – saving them from the cost of downloading the data from
Netflix server before passing it on to the viewers at home. By collecting end-user data
on how the physical location of the content affects the viewer’s experience, calculations
about the placement of data can be made to ensure an optimal service to as many
homes as possible. Data points such a delays due to buffering (rebuffer rate) and bitrate
(which affects the picture quality – if you’re watching a film on Netflix that suddenly
seems to switch from razor-sharp HD to a blurry mess, you’ve experienced a bitrate
drop) are collected to inform this analysis. Netflix has used Big Data and analytics to
Page 97
BIG DATA MARKETING
position itself as the clear leader of the pack. It has done this by taking on other
distribution and production networks at their own game, and trumping them through
innovative and constantly evolving use of data. It faces competition now and into the
future of course – most notably Amazon, which purchased UK-based Netflix rival
Lovefilm in 2011, acquiring its user data and melding it with their own advanced data
infrastructure and analysis platforms. Will Amazon – which pioneered the art of
recommendations before Netflix even existed – unseat Netflix from its perch as king of
the streaming content providers? Not to mention that Apple is about to launch it’s new
AppleTV service to compete in this space. Time will tell – but the race to develop more
accurate and insightful analytic strategies will be a key decider.
4.2 AMAZON
Link: https://www.forbes.com/sites/jonmarkman/2017/06/05/amazon-using-ai-big-
data-to-accelerate-profits/#ca9feae6d55f
It brought personal digital assistants into our kitchens with Echo, a connected speaker.
Now it can see, so Amazon wants to come into our bedrooms to help us choose more
flattering outfits.
The new Echo Look appliance features a depth-sensing camera, built-in lighting and Style
Check software that uses the latest advances in machine learning.
It’s a big jump for the technology. It could be bigger for Amazon’s bottom line.
The online retailer claims the algorithms are supplemented by fashion specialists. When
you tell Echo Look to take a picture or video, it will send the image to the cloud, store it,
perform data analytics, then offer recommendations based on current trends and what
looks good on you.
Page 98
BIG DATA MARKETING
In theory, it’s cool. It’s kind of like having your own personal stylist. In reality, Amazon is
dictating what is cool. Telling you what its algorithms have been trained to think looks
good. And all of it can be yours for one low price. Just say the word.
The market is already big. Statista forecasts online sales of footwear, apparel and
accessories will reach $90.34 billion in 2020. That is up from 2017 estimates of $69.82
billion.
Still, by Amazon standards, cracking the code has been a struggle thus far. It would
dearly love to be a bigger player in the “fast fashion” business. That’s the lucrative part
of the business that pushes fashion-show trends into retail stores at staggering speeds.
It’s trendy stuff with lots of repeat customers. That should be Amazon’s wheelhouse.
But it has not really worked out that way. Buying clothing online is hit and miss.
Customers want to know how garments fit, how they feel. Echo Look will help. It will
know everything about you, especially your size and your preferences. It will live in the
cloud, so you can take it on the road. And it’s always learning.
Alexa, take a selfie. What do you think? Does this make me look fat?
Making it easier for customers to buy things they love has been the basis of Amazon’s
recommendation engine.
In many ways, it’s a natural evolution. Since inception, Amazon has been using artificial
intelligence to make it easier for customers to buy things they love. It’s the basis of its
recommendation engine. Using AI and the copious amounts of data most people freely
offer online just makes sense.
It’s also the clearest way to monetize that data. It’s the best business model. This is
important because other technology firms are in the game, too.
Recently, Alphabet announced its connected speaker, Google Home, has the ability to
recognize different voices.
Page 99
BIG DATA MARKETING
A family could use a single Home device to access multiple Google accounts. Dad can
ask for his commute time to work. Mom can access her calendar. The kids can listen to
their own Spotify playlists.
AI neural networks recognize who is talking and the context of their words. Then they
return information based on the account holder. It’s heady stuff.
And Facebook is getting close to a general release of its M digital assistant. M is an AI-
and-human-trained hybrid, so it can do some amazing things. M lives in Messenger chat
and can send money (naturally), offer suggestions for restaurants you might like based
on location, hail a ride, set reminders and more.
M has complete access to the user’s Facebook account. It knows your friends, your
pictures, what dates are important to you and what you like. It even knows what your
friends like.
M, Google Home and Alexa are revolutionary because they comb through massive
amounts of data people freely offer. And, as much as we claim to value privacy as a
society, our personal actions suggest otherwise.
It seems absurd that people would bring a camera and microphone, connected to an
online store, into their bedroom. The potential for abuse is unmistakable. Yet these are
the times we live in.
And that is why the investment story for AI is just getting started. We are at the
beginning of what AI will do for corporate profits — right along with how it will help
make life more friction-free.
As I’ve been saying for a few years, the best way for investors to take advantage of the
explosion of AI is to own the software giants — Amazon, Facebook and Alphabet.
Link: https://www.youtube.com/watch?v=9X_fP4pPWPw
Page 100
BIG DATA MARKETING
4.3 NIKE
Article: What Nike’s Analytics Platform Buy Says About the Future of Brand-Consumer
Relationships
Link: https://www.adweek.com/digital/what-nikes-analytics-platform-buy-says-about-
the-future-of-brand-consumer-relationships/
When Nike announced its acquisition of analytics platform Zodiac in March, it spoke in
broad terms about its plans to build closer relationships with consumers.
Nike and its agency of record declined comment, but bringing a predictive analytics firm
in-house will undoubtedly help it crunch all the data it collects as it rolls out connected
products and experiences and will also help Nike make predictions about those
consumers.
A Nike rep said the acquisition enhances the brand’s ability to serve consumers one-to-
one as a result of a deeper understanding of their goals and needs, and Zodiac also
brings with it a team of data science talent.
Michael Horn, managing director of data science at digital marketing agency Huge, said
the deal comes amid a “ferocious hiring climate” for top data science talent.
“So this is just as likely to be an acqui-hire of staff, plus some IP, to accelerate Nike’s
ambitions,” he said. “It’s hard enough to a brand as visible and admired as Nike to build
a data science team—let alone hundreds of other brands which are struggling to attract
and pay for transformative data talent.”
As of March 29, Nike has more than 300 open jobs listed under the keyword “data.”
Page 101
BIG DATA MARKETING
“I think this is a trend you’ll see a lot of brands with strong direct-to-consumer initiatives
taking—building up in-house data science teams as well as strengthening the customer
data platforms, not just because retailers tend to not share this with brands, but
moreover the brands need the deep data insights for more accurate personalization of
their products and services,” said Manolo Almagro, managing director at technology
consultancy Q Division.
Horn pointed out that Nike has long recognized an opportunity to differentiate itself at
the intersection of wearables, IoT and quantified health as it puts sensors on everything
from basketball courts to shoes to uniforms.
“That data streams in by the fraction of a second and can be used to predict the
performance of fitness enthusiasts, professional athletes or even entire sports teams,”
Horn said. “Nike experiences like Nike+ and the Run Club [which Huge helped build] are
the future of customer relationships for the brand and drive investments like this. And
their competitors [like Under Armour] are doing the same thing.”
In 2015, Under Armour acquired diet and exercise app MyFitnessPal for $457 million.
Horn said the app had 80 million users at the time but would have also brought plenty
of data infrastructure and talent with it, which was surely part of Under Armour’s
strategy. (Under Armour recently disclosed a breach in which the data—including email
addresses, user names and passwords—of 150 million MyFitnessPal users was exposed.)
In a release, Zodiac CEO Artem Mariychin said the analytics firm is excited to help power
Nike’s Consumer Direct Offense, which according to a 2017 Nike announcement, is a
reorganization that included the creation of the Nike Direct organization, which united
Nike.com, direct-to-consumer retail and Nike+ digital products, and will focus on 12
cities, including New York, London, Shanghai, Beijing, Los Angeles, Tokyo, Paris, Berlin,
Mexico City, Barcelona, Seoul and Milan. Nike projects that these cities and the
countries they are in will account for over 80 percent of its growth through 2020.
Page 102
BIG DATA MARKETING
Nike features by collecting and analyzing members’ shopping behavior, activity and
feedback across digital and physical properties. Rewards include access to Nike experts
on demand who provide personalized recommendations, an exclusive shop with
members-only products and options to personalize products, reserved products and
members-only shopping events.
Nike expanded NikePlus Unlocks in February 2018 to include access to music, guided
meditation and fitness classes based on members’ interests and goals.
“The more you use Nike.com and the Nike, NTC or NRC apps, the better Nike can get to
know you and serve you with promotions tailored to your interests and goals,” Nike said
in a release.
And, of course, the more consumers interact with Nike properties, the more data Nike
collects.
Ikea Family
Link: https://www.ikea.com/gb/en/ikea-family/
Mastercard
Link: https://www.mastercard.co.uk/en-gb/issuers/products-and-solutions/grow-
manage-your-business/loyalty-solutions/pay-with-rewards.html
Pizza Hut
Link: http://strategyonline.ca/2018/03/28/pizza-hut-introduces-hut-rewards/
My Starbucks Rewards
Link: https://www.starbucks.com/card/rewards/rewards-program-ts-and-cs
Page 103
BIG DATA MARKETING
5. BIBLIOGRAPHY
5.1 BOOKS
Mª Ángeles Herández Dauder, José María Estrade y David Jordán Soro. Marketing
Digital. Mobile marketing, SEO y analítica web. Edición 2017. Social Business.
5.2 WEBSITES
Dirk Breeuwer. Sobre los elementos clave para una estrategia de big data marketing.
https://www.inboundcycle.com/blog-de-inbound-marketing/big-data-marketing-los-3-
componentes-esenciales-para-una-estrategia-exitosa
Page 104
BIG DATA MARKETING
Pamela Chevez. Sobre los programas de fidelización. Omar Arce. Sobre la arquitectura
de un club de fidelización. http://ninjaclienting.com/arquitectura-club-fidelizacion-
rentable/
Page 105